Women's Fashion Product Recommendation System

A project for Capstone Project Subject and Graduation Thesis Subject of a fourth-year university student. (Not yet completed)

Time: Sept 2024 - present

Abstract

With the rapid growth of e-commerce, the demand for online fashion shopping, particularly in women's fashion, has significantly increased. However, the huge amount of information available has led to information overload, making it confusing for customers to make purchasing decisions. The aim of this project is to solve the challenge that customers are facing in making decisions on e-commerce platforms. The project consists of two main parts: collecting and analyzing data to evaluate current shopping trends and designing a recommendation system to assist customers in selecting suitable products.

Data Source

• All the data are collected by scraping real data from websites.
• The code for scraping data progress is in collect_analysis_data.ipynb.
• Collected data are stored in *.csv and *.txt formats.

Usage

To use this project, ensure you update file paths if you intend to import or load datasets using the provided code.

Technologies and techniques

• Web Scraping: Using Selenium for automated data extraction.
• Data Processing: Preprocessing techniques to clean and prepare data.
• Clustering: Implementing K-Means and DBSCAN for clustering product types.
• Natural Language Processing (NLP): Analyzing customer feedback.
• Visualization: Generating insights using data visualization.
• System Design: Designing the architecture for the recommendation system.

Notes

• The report included in this repository is for reference purposes only. Please do not edit or reuse it for any other purpose.
• All code and files are created by me. If you reuse any part of the code, please add appropriate citations.